Drought, conflict and children's undernutrition in Ethiopia 2000-2013: a meta-analysis

To estimate the prevalence of childhood wasting and to investigate the effects of drought and conflict on wasting in crisis-affected areas within Ethiopia.

We searched the Complex Emergency Database for nutrition surveys carried out in Ethiopia over the period 2000-2013. We extracted data on the prevalence of wasting (weight-for-height z-scores below -2) among children aged 6-59 months for areas of Ethiopia that had sufficient data available. Data on any conflict events (irrespective of magnitude or impact) and episodes of seasonal drought affecting the survey areas were extracted from publicly available data sources. Random-effects Bayesian meta-analysis was used to synthesize the evidence from 231 small-scale surveys.

From the total sample of 175 607 children analysed, the pooled number of children wasted was 21 709. The posterior median prevalence of wasting was 11.0% (95% credible interval, CrI: 10.3-11.7) over the 14-year period. Compared with areas unaffected by drought, the estimated prevalence of wasting was higher in areas affected by moderate levels of drought (posterior odds ratio, OR: 1.34; 95% CrI: 1.05-1.72) but similar in severe drought-affected areas (OR: 0.96; 95% CrI: 0.68-1.35). Although the pooled prevalence of wasting was higher in conflict-affected than unaffected areas, the difference was not plausible (OR: 1.02; 95% CrI: 0.82-1.26).

Despite an overall declining trend, a wasting problem persists among children in Ethiopia. Conflict events did not have a major impact on childhood wasting. Nutrition interventions should go beyond severe drought-prone areas to incorporate areas where moderate droughts occur.